Implementation of Organized Convection Parameterization in the Met Office Unified Model

Organized convection is one of the critical factors for modulating large-scale circulations through redistributing heat, moisture, and momentum. However, these redistributions associated with organized convection struggle to be represented in global models. For instance, organized convection releases more upper-tropospheric heating and lower-tropospheric cooling than discrete convection, but this top-heavy latent heating profile cannot be reproduced in most global climate models, partially because these models under-resolve the mesoscale overturning circulations and fail to represent their associated microphysical processes. To decrease these biases, Moncrieff et al. developed a Multiscale Coherent Structural Parameterization (MCSP) scheme for producing the top-heavy heating profile when coarse-grid environments are favorable for organized convection. However, favorable environments are not necessarily associated with organized convection, meaning that a deterministic approach might not fully capture the relationship between environmental conditions and appropriate triggering of the MCSP scheme.

To overcome this shortcoming, this study, as part of the MCS PRIME project, implements the MCSP scheme into the Met Office Unified Model (UM). As a first step we want to understand how MCSP behaves alongside with other schemes in this new model and later we will leverage the pre-existing UM stochastic scheme to trigger the MCSP scheme with a likelihood of organized convection in ensemble runs. In contrast to past studies, the MCSP scheme couples with an advanced mass-flux convection scheme (CoMorph) that explicitly parameterizes convective detrainment and entrainment and considers the interaction between shallow and deep clouds. Long-term global simulations and short-term ensemble runs focusing on convective cases are conducted to evaluate the scheme. Preliminary results show that the MCSP scheme can generate the top-heavy heating profile for given environmental conditions, impacting global precipitation and rain rate probability distributions. Future work will use stochastic schemes to represent organized convection probability as a function of environmental factors and potentially refine the MCSP impact on momentum and moisture budget through learning from data assimilation increments.